Pancreatic cancer is one of the most invasive human cancers that has become increasingly prevalent in recent years. It is estimated that by 2030, this cancer will be the second leading cause of death among cancers. Therefore, identifying sensitive and specific biomarkers for the early diagnosis and treatment of pancreatic cancer, as well as predicting its survival and prognosis, is crucial. High-throughput analysis can find gene expression differences and critical molecular pathways in normal and cancerous cases, leading to the development of biomarkers for better management of pancreatic cancer. In this study, a data set, including the expression data of 34,706 different genes in 82 different samples of pancreatic cancer and normal tissues, was analyzed in the TCGA database. Meanwhile, 2000 DEGs were found in cancer samples compared to normal tissues, considering the significance level of p < 0.05. All DEGs were down-expressed in pancreatic cancer in comparison to normal tissue. To have a deep understanding of the function of the DEGs, we performed enrichment analysis and PPI network analysis to screen the genes and pathways associated with pancreatic cancer that are more important in the development and progression of pancreatic cancer.
In our studies, the results of the principal component analysis showed that the top 20 DEGs might be potential diagnostic and prognostic biomarkers to improve pancreatic cancer treatment, including four genes encoding histone proteins called H1-4, H1-5, H4C3, and H4C2 and 16 genes encoding non-coding RNAs named RN7SL2, RN7SL3, RN7SL4P, RN7SKP80, SCARNA12, SCARNA10, SCARNA5, SCARNA7, SCARNA6, SCARNA21, SCARNA9, SCARNA13, SNORA73B, SNORA53, SNORA54, and SNORD17 (Table 1). Previous studies have revealed that histones, as chromatin-regenerating proteins, are essential in cancer pathogenesis. Histones undergo severe changes during cancer promotion/progression and may involve in causing the disease. Among the four genes encoding histones, histone H1 is a cancer promoter and a cancer biomarker in different malignancies [30–32]. The exact molecular function of the majority of non-coding RNAs found in the present study was unclear, and there are few articles about those non-coding RNAs. Several studies have shown that long-noncoding RNAs such as RN7SL2 and RN7SL4P are overexpressed in patients with multiple myeloma [33]. RN7SL2 is abundant in the cancer patient’s plasma [34]. In contrast, another report presented that the RN7SL3 is downregulated in hepatocellular carcinoma [35]. SNORA73 is a chromatin-associated snoRNA and is effective in genome stability [36, 37]. SNORA54 has been studied in many human cancers such as breast, melanoma, lymphoma, and myeloma. This snoRNA has upregulated in most cancer patients but is down-expressed in patients with melanoma [38, 39]. According to the literature, SNORD17 is overexpressed in cases of hepatocellular carcinoma, and its upregulation is usually associated with poor clinical outcomes [40, 41].
Unfortunately, no pancreatic cancer study has been performed on the found non-coding RNA expression and function. However, one study demonstrated that SCARNA6 is overexpressed in patients with autism spectrum disorders [42]. This finding can be significant due to the close relationship between gene expression in the pancreas and neural tissues. SCARNA7 is correlated with many cancers such as breast, prostate, and non-small cell lung cancers. This SCARNA is usually upregulated in breast cancer, and it is associated with poor prognosis [43–45]. The findings of the other study revealed that SCARNA9 was significantly overexpressed in colon cancer. In contrast, another study suggested that downregulation of SCARNA9 is negatively associated with endometrial cancer [46, 47]. Numerous studies have investigated the expression of SCARNA10 in liver fibrosis and hepatocellular carcinoma. These studies showed that the expression of SCARNA10 increased, and is usually associated with the physio-pathological features of these diseases. Hence, this SCARNA has been introduced as a diagnostic biomarker and therapeutic target in liver fibrosis and hepatocellular carcinoma. Silencing of SCARNA10 gene in hepatocytes has displayed down-expression of TGFβ, TGFβRI, SMAD2, SMAD3, and KLF6 [48–51]. SCARNA13 is highly expressed in hepatocellular carcinoma, and it is involved in tumorigenesis and metastasis [52–54].
GO analysis of the top 20 DEGs in our study showed that those are mainly enriched in the pathways associated with negative regulation of gene silencing, negative regulation of chromatin organization, negative regulation of chromatin silencing, nucleosome positioning, regulation of chromatin silencing, DNA packaging, negative regulation of DNA metabolic process, regulation of DNA recombination, chromosome condensation, negative regulation of DNA recombination, positive regulation of gene expression, epigenetic regulation, chromatin assembly, nucleosome organization, positive regulation of histone H3-K9 methylation, the establishment of protein localization to chromatin, Histone H3-K9 methylation, protein localization to chromosome, protein localization to chromatin and positive regulation of histone methylation/DNA binding (Table 2). Interestingly, among the 20 DEGs, only two genes, H1-4 and H1-5, were identified as influential genes in GO analysis.
The KEGG analysis of the top 20 DEGs demonstrated a relationship between pancreatic cancers and other diseases such as systemic lupus erythematosus, alcoholism, neutrophil extracellular trap formation, and viral carcinogenesis, due to the function of H4C2 and H4C3 genes. Other studies have proved that alcoholism (consumption of high amounts of alcohol) is one of the critical risk factors in the progression and development of pancreatic cancer, especially in patients with Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations [55–58]. There have been many reports on the effect of neutrophil extracellular trap (NET) formation in pancreatic cancer, but its exact role in the development of pancreatic cancer is still unknown. At present, only one article has pointed to the anti-cancer effects of NET, still, most studies have emphasized the function of NET formation in symptom exacerbation, resistance to immunotherapy, and induction of migration and invasion in pancreatic cancer cells. The NET formation has even been suggested to be involved in predicting the survival of pancreatic cancer patients after surgery [59–62]. There is ample evidence linking systemic lupus erythematosus to the risk of developing various cancers. In a meta-analysis study, systemic lupus erythematosus was associated with an increased risk of pancreatic cancer [63]. But in another study, no significant relationship was found between the two diseases [64]. In line with our results, other studies show the role of viral infections in pancreatic carcinogenesis including the SARS family of coronaviruses and the hepatitis family (B and C). Certainly, careful monitoring of patients with these diseases may help in the early diagnosis of pancreatic cancer and predict the prognosis of these patients [65–67].
In the last part of this study, PPI network analysis was performed for the top 100 DEGs. As described in the results (Fig. 7A), 14 nodes were identified with these DEGs. Eleven nodes with a degree of connectivity equal to 13 were selected as hub genes. Interestingly, these hub genes were histone-encoding genes, include H4C3, H1-4, H4C2, H1-5, H4C5, H4C4, H2AC20, H2AC14, H2BC13, H2AC17, H2BC3, H1-3, H2AC12, H2AC21, H2BC17, H2AC4, H4C8, H2BC6, H3C7, H3C11, H4C1, H4C6 and H4C13. These results indicated the role of histones in the development of pancreatic cancer. Numerous reports suggest that histone gene expression profiles in many cancer types such as breast, lung, prostate, kidney, and pancreas may play a role in pancreatic cancer prognosis. For instance, the expression of histone H1.3 in pancreatic cancer patients can predict the clinical outcome after pancreatic surgery. Therefore, H1.3 was identified as one of the prognostic biomarkers in pancreatic cancer [30, 32, 68, 69]. Finally, studies on histones and non-coding RNAs should be performed to determine their role and function in pancreatic cancer.